]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/kaltura.py
41c1f3d96be288957cbaba2c21e3477a2969b24d
[youtubedl] / youtube_dl / extractor / kaltura.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5 import base64
6
7 from .common import InfoExtractor
8 from ..compat import (
9 compat_urlparse,
10 compat_parse_qs,
11 )
12 from ..utils import (
13 clean_html,
14 ExtractorError,
15 int_or_none,
16 unsmuggle_url,
17 smuggle_url,
18 )
19
20
21 class KalturaIE(InfoExtractor):
22 _VALID_URL = r'''(?x)
23 (?:
24 kaltura:(?P<partner_id>\d+):(?P<id>[0-9a-z_]+)|
25 https?://
26 (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com(?::\d+)?/
27 (?:
28 (?:
29 # flash player
30 index\.php/(?:kwidget|extwidget/preview)|
31 # html5 player
32 html5/html5lib/[^/]+/mwEmbedFrame\.php
33 )
34 )(?:/(?P<path>[^?]+))?(?:\?(?P<query>.*))?
35 )
36 '''
37 _SERVICE_URL = 'http://cdnapi.kaltura.com'
38 _SERVICE_BASE = '/api_v3/index.php'
39 # See https://github.com/kaltura/server/blob/master/plugins/content/caption/base/lib/model/enums/CaptionType.php
40 _CAPTION_TYPES = {
41 1: 'srt',
42 2: 'ttml',
43 3: 'vtt',
44 }
45 _TESTS = [
46 {
47 'url': 'kaltura:269692:1_1jc2y3e4',
48 'md5': '3adcbdb3dcc02d647539e53f284ba171',
49 'info_dict': {
50 'id': '1_1jc2y3e4',
51 'ext': 'mp4',
52 'title': 'Straight from the Heart',
53 'upload_date': '20131219',
54 'uploader_id': 'mlundberg@wolfgangsvault.com',
55 'description': 'The Allman Brothers Band, 12/16/1981',
56 'thumbnail': 're:^https?://.*/thumbnail/.*',
57 'timestamp': int,
58 },
59 },
60 {
61 'url': 'http://www.kaltura.com/index.php/kwidget/cache_st/1300318621/wid/_269692/uiconf_id/3873291/entry_id/1_1jc2y3e4',
62 'only_matching': True,
63 },
64 {
65 'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3',
66 'only_matching': True,
67 },
68 {
69 'url': 'https://cdnapisec.kaltura.com/html5/html5lib/v2.30.2/mwEmbedFrame.php/p/1337/uiconf_id/20540612/entry_id/1_sf5ovm7u?wid=_243342',
70 'only_matching': True,
71 },
72 {
73 # video with subtitles
74 'url': 'kaltura:111032:1_cw786r8q',
75 'only_matching': True,
76 },
77 {
78 # video with ttml subtitles (no fileExt)
79 'url': 'kaltura:1926081:0_l5ye1133',
80 'info_dict': {
81 'id': '0_l5ye1133',
82 'ext': 'mp4',
83 'title': 'What Can You Do With Python?',
84 'upload_date': '20160221',
85 'uploader_id': 'stork',
86 'thumbnail': 're:^https?://.*/thumbnail/.*',
87 'timestamp': int,
88 'subtitles': {
89 'en': [{
90 'ext': 'ttml',
91 }],
92 },
93 },
94 'skip': 'Gone. Maybe https://www.safaribooksonline.com/library/tutorials/introduction-to-python-anon/3469/',
95 'params': {
96 'skip_download': True,
97 },
98 },
99 {
100 'url': 'https://www.kaltura.com/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto',
101 'only_matching': True,
102 },
103 {
104 'url': 'https://www.kaltura.com:443/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto',
105 'only_matching': True,
106 }
107 ]
108
109 @staticmethod
110 def _extract_url(webpage):
111 # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site
112 mobj = (
113 re.search(
114 r"""(?xs)
115 kWidget\.(?:thumb)?[Ee]mbed\(
116 \{.*?
117 (?P<q1>['"])wid(?P=q1)\s*:\s*
118 (?P<q2>['"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
119 (?P<q3>['"])entry_?[Ii]d(?P=q3)\s*:\s*
120 (?P<q4>['"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
121 """, webpage) or
122 re.search(
123 r'''(?xs)
124 (?P<q1>["'])
125 (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
126 (?P=q1).*?
127 (?:
128 entry_?[Ii]d|
129 (?P<q2>["'])entry_?[Ii]d(?P=q2)
130 )\s*:\s*
131 (?P<q3>["'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
132 ''', webpage) or
133 re.search(
134 r'''(?xs)
135 <iframe[^>]+src=(?P<q1>["'])
136 (?:https?:)?//(?:www\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)
137 (?:(?!(?P=q1)).)*
138 [?&]entry_id=(?P<id>(?:(?!(?P=q1))[^&])+)
139 (?P=q1)
140 ''', webpage)
141 )
142 if mobj:
143 embed_info = mobj.groupdict()
144 url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
145 escaped_pid = re.escape(embed_info['partner_id'])
146 service_url = re.search(
147 r'<script[^>]+src=["\']((?:https?:)?//.+?)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
148 webpage)
149 if service_url:
150 url = smuggle_url(url, {'service_url': service_url.group(1)})
151 return url
152
153 def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
154 params = actions[0]
155 if len(actions) > 1:
156 for i, a in enumerate(actions[1:], start=1):
157 for k, v in a.items():
158 params['%d:%s' % (i, k)] = v
159
160 data = self._download_json(
161 (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
162 video_id, query=params, *args, **kwargs)
163
164 status = data if len(actions) == 1 else data[0]
165 if status.get('objectType') == 'KalturaAPIException':
166 raise ExtractorError(
167 '%s said: %s' % (self.IE_NAME, status['message']))
168
169 return data
170
171 def _get_video_info(self, video_id, partner_id, service_url=None):
172 actions = [
173 {
174 'action': 'null',
175 'apiVersion': '3.1.5',
176 'clientTag': 'kdp:v3.8.5',
177 'format': 1, # JSON, 2 = XML, 3 = PHP
178 'service': 'multirequest',
179 },
180 {
181 'expiry': 86400,
182 'service': 'session',
183 'action': 'startWidgetSession',
184 'widgetId': '_%s' % partner_id,
185 },
186 {
187 'action': 'get',
188 'entryId': video_id,
189 'service': 'baseentry',
190 'ks': '{1:result:ks}',
191 },
192 {
193 'action': 'getbyentryid',
194 'entryId': video_id,
195 'service': 'flavorAsset',
196 'ks': '{1:result:ks}',
197 },
198 {
199 'action': 'list',
200 'filter:entryIdEqual': video_id,
201 'service': 'caption_captionasset',
202 'ks': '{1:result:ks}',
203 },
204 ]
205 return self._kaltura_api_call(
206 video_id, actions, service_url, note='Downloading video info JSON')
207
208 def _real_extract(self, url):
209 url, smuggled_data = unsmuggle_url(url, {})
210
211 mobj = re.match(self._VALID_URL, url)
212 partner_id, entry_id = mobj.group('partner_id', 'id')
213 ks = None
214 captions = None
215 if partner_id and entry_id:
216 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
217 else:
218 path, query = mobj.group('path', 'query')
219 if not path and not query:
220 raise ExtractorError('Invalid URL', expected=True)
221 params = {}
222 if query:
223 params = compat_parse_qs(query)
224 if path:
225 splitted_path = path.split('/')
226 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
227 if 'wid' in params:
228 partner_id = params['wid'][0][1:]
229 elif 'p' in params:
230 partner_id = params['p'][0]
231 elif 'partner_id' in params:
232 partner_id = params['partner_id'][0]
233 else:
234 raise ExtractorError('Invalid URL', expected=True)
235 if 'entry_id' in params:
236 entry_id = params['entry_id'][0]
237 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
238 elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
239 reference_id = params['flashvars[referenceId]'][0]
240 webpage = self._download_webpage(url, reference_id)
241 entry_data = self._parse_json(self._search_regex(
242 r'window\.kalturaIframePackageData\s*=\s*({.*});',
243 webpage, 'kalturaIframePackageData'),
244 reference_id)['entryResult']
245 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
246 entry_id = info['id']
247 # Unfortunately, data returned in kalturaIframePackageData lacks
248 # captions so we will try requesting the complete data using
249 # regular approach since we now know the entry_id
250 try:
251 _, info, flavor_assets, captions = self._get_video_info(
252 entry_id, partner_id)
253 except ExtractorError:
254 # Regular scenario failed but we already have everything
255 # extracted apart from captions and can process at least
256 # with this
257 pass
258 else:
259 raise ExtractorError('Invalid URL', expected=True)
260 ks = params.get('flashvars[ks]', [None])[0]
261
262 source_url = smuggled_data.get('source_url')
263 if source_url:
264 referrer = base64.b64encode(
265 '://'.join(compat_urlparse.urlparse(source_url)[:2])
266 .encode('utf-8')).decode('utf-8')
267 else:
268 referrer = None
269
270 def sign_url(unsigned_url):
271 if ks:
272 unsigned_url += '/ks/%s' % ks
273 if referrer:
274 unsigned_url += '?referrer=%s' % referrer
275 return unsigned_url
276
277 data_url = info['dataUrl']
278 if '/flvclipper/' in data_url:
279 data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
280
281 formats = []
282 for f in flavor_assets:
283 # Continue if asset is not ready
284 if f.get('status') != 2:
285 continue
286 # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
287 # skip for now.
288 if f.get('fileExt') == 'chun':
289 continue
290 if not f.get('fileExt'):
291 # QT indicates QuickTime; some videos have broken fileExt
292 if f.get('containerFormat') == 'qt':
293 f['fileExt'] = 'mov'
294 else:
295 f['fileExt'] = 'mp4'
296 video_url = sign_url(
297 '%s/flavorId/%s' % (data_url, f['id']))
298 # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
299 # -f mp4-56)
300 vcodec = 'none' if 'videoCodecId' not in f and f.get(
301 'frameRate') == 0 else f.get('videoCodecId')
302 formats.append({
303 'format_id': '%(fileExt)s-%(bitrate)s' % f,
304 'ext': f.get('fileExt'),
305 'tbr': int_or_none(f['bitrate']),
306 'fps': int_or_none(f.get('frameRate')),
307 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
308 'container': f.get('containerFormat'),
309 'vcodec': vcodec,
310 'height': int_or_none(f.get('height')),
311 'width': int_or_none(f.get('width')),
312 'url': video_url,
313 })
314 if '/playManifest/' in data_url:
315 m3u8_url = sign_url(data_url.replace(
316 'format/url', 'format/applehttp'))
317 formats.extend(self._extract_m3u8_formats(
318 m3u8_url, entry_id, 'mp4', 'm3u8_native',
319 m3u8_id='hls', fatal=False))
320
321 self._sort_formats(formats)
322
323 subtitles = {}
324 if captions:
325 for caption in captions.get('objects', []):
326 # Continue if caption is not ready
327 if f.get('status') != 2:
328 continue
329 if not caption.get('id'):
330 continue
331 caption_format = int_or_none(caption.get('format'))
332 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
333 'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
334 'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
335 })
336
337 return {
338 'id': entry_id,
339 'title': info['name'],
340 'formats': formats,
341 'subtitles': subtitles,
342 'description': clean_html(info.get('description')),
343 'thumbnail': info.get('thumbnailUrl'),
344 'duration': info.get('duration'),
345 'timestamp': info.get('createdAt'),
346 'uploader_id': info.get('userId') if info.get('userId') != 'None' else None,
347 'view_count': info.get('plays'),
348 }